OBJECTIVE: It is suggested that age at onset (AAO) of bipolar I disorder (BP-I) is decreasing. We tested for a birth-cohort effect on AAO using admixture analysis.
METHOD: A clinical sample of 3896 BP-I cases was analysed using two approaches: (i) in a subsample with untruncated AAO × birth year distribution (n = 1865), we compared the best-fitting model for the observed AAO in patients born ≤1960 and >1960, (ii) to control for potential confounders, two separate subsamples born ≤1960 and >1960 were matched for age at interview (n = 250), and a further admixture analysis was undertaken.
RESULTS: The two approaches indicated that the proportion of cases in the early AAO category was significantly greater in cases born >1960; manic onsets were also more frequent in the early onset BP-I cases born >1960.
CONCLUSION: The decrease in AAO of BP-I in recent birth-cohorts appears to be associated with an increase in the proportion of cases in the early onset subgroup; not with a decrease in the mean AAO in each putative subgroup. This could indicate temporal changes in exposure to risk factors for mania.